The basis of all studies of personality, including its origins, universality, and development, lies in the description of personality structure. The organization of traits provides a framework for systematic research on all aspects of trait psychology. Although the Five-Factor Model of personality is widely recognized as providing a necessary and reasonably sufficient classification of traits, some researchers have argued that the five factors are themselves organized by a higher, two-factor model, in which Extraversion and Openness form one factor, and Agreeableness and Conscientiousness vs. Neuroticism define the other. Such a structure is widely found in self-report studies, including our study of personality in 12- to 13-year-olds. But when a single method of measurement is used in single individuals, it is impossible to determine whether the two-factor structure represents the real covariation of traits (substance) or simply biases in measurement (artifact). In a collaborative study examining personality in twins and using both self-reports and informant ratings, we found much more consistent evidence for the artifact interpretation. It appears that the Five-Factor Model reflects the highest level of trait structure;higher levels are due to biases in measurement. Previous research in this Laboratory has examined the origins and accuracy of national character stereotypes. A recent study examined the Global Leadership and Organizational Behavior Effectiveness (GLOBE)scales to further the goals of interpretation of national character sterotypes. The Global Leadership and Organizational Behavior Effectiveness (GLOBE) Research Program is a large-scale study of organizational leadership in cultural context. Parts of the survey gathered in 62 nations were intended to characterize the culture as a whole in terms of typical practices and values. Four of the nine Societal Practices scales seem, directly or indirectly, to assess features of personality. These GLOBE scales ask for descriptions of typical personality traits that might be interpreted as judgments of national character. Ratings of national character reflect cultural identities and social dynamics, but previous research suggests that they are unrelated to the mean personality traits of the culture's members. Analyses at the culture level comparing GLOBE scales with aggregate assessed personality traits (n = 34) and with measures of perceived national character (n = 33) showed that these GLOBE scales are better construed as unfounded stereotypes than as actual depictions of the society members'personality traits. In a further study, we considered the prediction of other nations'views of Americans. Contrary to the evaluative hypothesis, which assumes that desirable traits will be attributed to Americans by friendly nations and undesirable traits by hostile nations, we found that both desirable and undesirable traits were attributed to Americans by citizens of most countries. A new multinational study includes assessments of stereotypes of age groups and genders as well as nations. Preliminary analyses suggest that (a) age is more important that gender or culture in determining perceptions of a group;(b) age stereotypes are generally accurate in the direction of age changes, but not the timing;and (c) the stereotype of the typical culture member can be approximated by summing across age- and gender-specific national stereotypes. These analyses will provide important new information on the conditions under which stereotypes are, or are not, accurate. Comparisons of personality profiles across cultures provide some of the strongest evidence of the accuracy of aggregate, or average, personality scores. But there are a number of different metrics by which profile agreement can be assessed. In a study using self-reports and observer ratings of 532 individuals, the ability of several different profile agreement statistics to distinguish matched (i.e., the self-report and observer rating concerned the same individual) from mismatched pairs of self-reports and observer ratings was assessed. Intraclass correlations proved the best, with simple Pearson correlations a close second. In comparing personality profiles across cultures, intraclass correlations appear to be the metric of choice.